Retroviruses are important systems for studying carcinogenesis. Both acute transforming retroviruses (carrying viral oncogenes) and non-acute retroviruses (lacking oncogenes) induce specific neoplasms with high frequencies and predictable time courses. As such, they provide ideal systems for studying the multi-step carcinogenic process. In this proposal, we will continue our experiments on leukemogenesis by Moloney murine leukemia virus (M-MuLV). Early events will be the major focus. An important tool will be a poorly leukemogenic enhancer variant, Mo+PyF101 M-MuLV. Comparisons of mice infected by this virus under conditions where it is poorly leukemogenic vs. wild-type infected mice will provide a powerful method to distinguish particular cells or changes relevant to leukemogenicity. Four aims will be pursued: 1) Initial events (1-10d) for leukemogenesis will be studied. Previous studies indicate that early bone marrow infection is important for M-MuLV leukemogenesis. We will identify the cell types infected in the bone marrow (stromal vs. hematopoietic). Comparisons of wt vs. Mo+PyF101 infection will allow us to determine which bone marrow cells are important in leukemogenesis. Comparisons between these two viruses will also allow us to identify cells responsible for delivery of infection from the skin to the bone marrow, since Mo+PyF101 is poorly leukemogenic when infected SC, but efficiently leukemogenic when infected IP. 2) Events during intermediate preleukemic times (4-8 weeks) will be studied. Preleukemic splenomegaly will be studied, and the question of stromal cell infection will be addressed. Expression of cytokines in the preleukemic spleens will also be studied. Accelerated thymic atrophy will also be studied, with attention to cell types infected, and the potential roles of MCF recombinants and apoptosis. 3) Studies on MCF recombinant propagation and formation will be carried out. We will attempt to identify the site of MCF formation in vivo, with particular attention to the bone marrow. RT-PCR assays will detect transcription of endogenous polytropic transcripts as well as MCF recombinants. Cells that propagate MCFs in various tissues will be identified by time infections with M-MCF or by a replication-defective beta-gal/MCF vector. 4) A novel MuLV from China (SRS) that induces tumors of multiple lineages will be studied. The DNA sequence of an infectious molecular clone will be determined, chimeras with M-MuLV will be used to identify the regions that control the disease specificity, preleukemic states will studied, and we will investigate if multi-potential hematopoietic cells are transformed.